You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In addition to showing the amount of time, a given commit haven't been integrated, one could also measure how much (as in changed lines of code) change is laying on a given branch:
Thanks @sofusalbertsen I also think that the size of a branch would be interesting, with multiple measures.
Using diff might be a better choice then log? Are we interested in the size of the potential merge or are we interested in the number of times lines are changed, even the same line?
We should also be aware that to many stats will make it harder to find what is important for the user.
I think still that age is more important then "size".
I see this as an army knife of utills that can give you insights, where each metric should have a reason/description about what this metric can tell you.
But maybe this is scope creep for your repo's intention?
No, I don't think it's scope creep. I thinking it's a good idea and that we need to add it as new command or a flag on a the existing command.
This could also be a good oppertunity to make all the different endpoints "true" git commands by giving them the correct name. Like "git-branch-diffsize" or other?!
In addition to showing the amount of time, a given commit haven't been integrated, one could also measure how much (as in changed lines of code) change is laying on a given branch:
git log --numstat --pretty="%H" commit1..commit2 | awk 'NF==3 {plus+=$1; minus+=$2} END {printf("+%d, -%d\n", plus, minus)}'
The text was updated successfully, but these errors were encountered: